The goal of this project is to implement simulations of activation of Ca 2+/calmodulin-dependent vrotein kinase II (CaMKII) within the models of calcium dynamics in glutamatergic spines created in Project 1. We will test the accuracy of our understanding of spine biochemistry by designing experiments to verify or falsify predictions made from the simulations of activation of CaMKII. The project includes four specific aims, some of which will be pursued in parallel. First, we will design an accurate kinetic model of activation of the CaMKII holoenzyme at levels of calcium and at ratios of calmodulin to CaMKII that are likely to occur in spines. We will develop theoretical kinetic models, determine boundaries on appropriate kinetic constants, and fit the kinetic models to experiments with purified CaMKII. Second, we will introduce representations of holoenzymes of CaMKII, programmed with appropriate activation kinetics, into the models of calcium dynamics in spines created in Project i and carry out simulations of activation of CaMKII. We will compare simulations in which kinetic parameters are varied within known or likely physiological boundaries. We will compare our results to those obtained with more traditional finite element methods. Third, We will test the importance, in our models, of calmodulin location and concentration, phosphatase location and concentration, and also of other parameters that control calcium availability. Fourth, in collaboration with Karel Svoboda (project 4), we will design experiments to test whether simulations of CaMKII activation match experimental observations, and accurately predict behavior of CaMKII in spines under a variety of physiological conditions. Our experimental systems will include rat hippocampal slices, and cultured hippocampal neurons. The second and third aims will require extensive collaboration with the Core and with Project 1. We will work closely with the Sejnowski group to implement additions to the MCell software that are necessary for representing CaMKII. The models developed in Project 1, with input from Projects 2 and 4, will form the basis for models constructed in this Project.